Simulation of sedimentation rates using the SWAT model A case study of the Tarbela Dam, Upper Indus Basin Ahammad, H. I 1., Crosato A 2,3., Maskey, S 2., Masih, I 2,*, McClain, M 2. 1 Punjab Irrigation Department, Pakistan 2 IHE Delft Institute for Water Education, Westvest 7, 2611 AX, Delft, the Netherlands 3 Deltares, Boussinesqweg 1, 2629 HV Delft, the Netherlands. Corresponding Author: I. Masih (i.masih@un-ihe.org).
Outline of the Presentation Introduction Problem statement & Research objectives Methodology Selection of climate data (precipitation and temperature) Model setup for calibration and validation of hydrology and sediment Results of Calibration and Validation. Land use changes scenarios Conclusion Limitations Recommendations
Problem statement The main problems in the Upper Indus Basin are described as under: Soil erosion and sediment transport rate Unknown contribution of sediment from subbasins of the catchment Impact of Land use/cover (LULC) changes on discharge and sediment yield Optimization of Billion Tree Tsunami project.
Research Questions Based on above issues, the questions requiring further research are listed below: What are the current sediment inputs to Tarbela Reservoir? What are current contributions of the major sub-basins in terms of sediment production in the Upper Indus Basin? What will be the effects of the Billion Tree Tsunami project on the water and sediment inputs to the Tarbela Reservoir? Can the Billion Tree Tsunami project be optimized by assessing priority areas? Research Objectives This research will focus on change of land use/cover of the basin and its impacts on the sediment yield of basin at the sub basin scale.
Study area The Indus River originates from Tibetan Plateau of china. The Indus River is a trans-boundary river which originates from China and passes through India and then Pakistan having length 1126 km upto Tarbela Reservoir. The area of Upper Indus Basin upto Tarbela dam is 169,333 km 2 The Tarbela Reservoir is the downstream boundary of the Upper Indus Basin.
Topography The elevation in the Upper Indus Basin varies from 550 m to 8200 m (meter above sea level)
Methodology
Catchment delineation Model No. of sub basins- 23 Sub basin map
Land Use map Source: Land use map by Cheema (2010)
Soil map Source: FAO soil map used which is freely available on water base project website and having spatial resolution of 1 km x 1 km. Slope classification - 3 class 0-8, 8-30, and 30-99 percent of slope
Precipitation data Parameter TRMM PERSIANN-CDR CFSR mm/year mm/year mm/year Overall mean 298.2 415 724 STDEV 75.1 94.1 328 Min 180.1 207.6 235 Max 462.5 631.3 1529 Temperature data Observed + CFSR (Climate forecast system re analysis)
Model Setup for calibration & validation Discharge data obtained from WAPDA
Calibration parameters for hydrology
Results of Calibration and Validation for runoff Monthly time scale: S.No. Station name Station code Sub-basin outlet no. Calibration Validation R 2 NSE PBIAS R 2 NSE PBIAS 1 Kharmong - 10 0.80 0.79 +5.70 0.70 0.55-0.246 2 Pertab Bridge 3574760 4 0.84 0.80-4.48 0.85 0.82-10.4 3 Besham Qila 34729801 9 0.85 0.83 +11.3 0.84 0.81 +15.3 Daily time scale: S.No. Station name Station code Sub-basin outlet no. Calibration Validation R 2 NSE PBIAS R 2 NSE PBIAS 1 Besham Qila 34729801 9 0.76 0.70 +23 0.75 0.71 +14.4
Calibration at Besham Qila (Final Outlet station) Observed Simulated Average:2367.45 2040.7 R 2 =0.85 NSE=0.83 PBIAS= +11.3 10000 9000 Monthaly Calibration at Besham Qila Besham Qila 8000 observed simulated Discharge (m3/sec) 7000 6000 5000 4000 3000 2000 1000 0 2000/1 2000/3 2000/5 2000/7 2000/9 2000/11 2001/1 2001/3 2001/5 2001/7 2001/9 2001/11 2002/1 2002/3 2002/5 2002/7 2002/9 2002/11 2003/1 2003/3 2003/5 2003/7 2003/9 2003/11 2004/1 2004/3 2004/5 2004/7 2004/9 2004/11 2005/1 2005/3 2005/5 2005/7 2005/9 2005/11 2006/1 2006/3 2006/5 2006/7 2006/9 2006/11 time
Validation at Besham Qila Observed Simulated Average:2457 2081 R 2 =0.84 NSE=0.81 PBIAS= 15.3 14000 Monthly validation at Besham Qila Besham Qila Discharhe (m3/sec) 12000 10000 8000 6000 4000 observed simulated 2000 0 2007/1 2007/3 2007/5 2007/7 2007/9 2007/11 2008/1 2008/3 2008/5 2008/7 2008/9 2008/11 2009/1 2009/3 2009/5 2009/7 2009/9 2009/11 2010/1 2010/3 2010/5 2010/7 2010/9 2010/11 2011/1 2011/3 2011/5 2011/7 2011/9 2011/11 2012/1 2012/3 2012/5 2012/7 2012/9 2012/11 time
Calibration & validation at Besham Qila (Daily time scale)
Calibration parameters for sediment
Results of Calibration and Validation for sediment on Monthly scale S.No. Station name Station code Sub-basin outlet no. Calibration Validation R 2 NSE PBIAS R 2 NSE PBIAS 1 Besham Qila 34729801 23 0.86 0.84-12.4 0.88 0.85 +1.14 Milliom m3/yaer 800.0 700.0 600.0 500.0 400.0 300.0 200.0 100.0 0.0 Cumulative sediment yield at Besham Qila period (2000-2010) observed simulated 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Year
Sediment calibration at Besham Qila Observed Average: 419618 tonnes Simulated 470558 tonnes NSE=0.84 PBIAS= - 12.4 Observed v/s Simulated corelation 4500000 Simulated 4000000 3500000 3000000 2500000 2000000 1500000 1000000 500000 0 R² = 0.855 0 1000000 2000000 3000000 4000000 5000000 Observed
Validation for Sediment at Besham Qila Observed Average: 573121 tonnes Simulated 566581 tonnes NSE=0.85 PBIAS= + 1.14 Observed v/s Simulated corelation 7000000 6000000 5000000 R² = 0.8842 Simulated 4000000 3000000 2000000 1000000 0 0 2000000 4000000 6000000 8000000 10000000 Observed
Total sediment yield comparison for Calibration & validation
Land use Land cover change scenarios i) 20% savanna (SAVD) replaced by forest (FRDA)
Land use Land cover change scenarios ii) 40% savanna (SAVD) replaced by forest (FRDA)
Sediment yield map of UIB Middle part of the catchment producing more sediment due to steep slopes with combination of lithosols.
Conclusion The results shows Billion Tree Tsunami project does not have significant impact on sediment yield of the basin even if 40% of the area under savana is replaced by forest. Morris (2014) based on observed data computed an average sediment yield of about 123 x 10 6 m 3 per year at Besham Qila. However, the calibrated/validated SWAT model estimate is about 96 x 10 6 m 3 per year. The most soil erosion prone sub basins as shown in sediment yield map, which will be useful for practicing engineers for watershed management. The middle part of the UIB emerges as the most erosion prone area.
Limitations of the study The meteorological stations are scarcely gauged within the administrative boundary of Pakistan and do not reflect the true representation of the catchment. Because of these limitations, complete meteorological data could not be obtained and it was not possible to develop the model on the basis of observed weather stations data. There were lack of sediment data availability at Kharmong and Partab Bridge gauging stations. Therefore, for sediment load calibration was not carried out but only was calibrated for runoff at these stations. Available observed meteorological data are found inconsistent due to large number of missing records which prohibited an acceptable degree of correlation analysis with satellite meteorological data. The results of model on north eastern catchments with area under glaciers were quite uncertain, due to SWAT s inability to adequately simulate glacier melt process.
Recommendations Future studies should also investigate the impact of other land use changes like urbanization, agriculture and deforestation. The future LULC change scenarios needs to be developed by considering socioeconomic trends in the catchment. Future studies should also look at erosions contribution due to landslides and earthquakes. The impact of proposed reservoirs upstream of Tarbela on sediment rates could be done using modelling approaches. The field level studies should be done on high sediment producing areas (middle part of the UIB) to formulate a sustainable sediment management.
Thanks! Questions?